AI leaders are divided over how long it will take to achieve AGI, machines that can reason like humans.
Google DeepMind CEO Demis Hassabis says “AI agents” are the next stage of AI after chatbots.
These systems will be able to plan, act, reason and use tools, among other characteristics, he said.
Experts disagree about what it takes to achieve artificial general intelligence, a still hypothetical form of robotic intelligence that mimics human abilities.
According to Google DeepMind CEO and recent Nobel Prize winner Demis Hassabis, there’s no secret formula to get there.
At the Times Tech Summit earlier this month, Hassabis said the next AI after chatbots like ChatGPT, which passively answer questions, summarize text, and conduct research, will be agent-based with the following characteristics and skills: He said it would be a system.
Under planning. These systems can help you plan your trip and book tickets ahead of time.
acting. They need to be able to take action in the real world.
Reasoning. They should think through the problem. Hassabis pointed to DeepMind’s AI, AlphaGo, which was the first to beat humans at Go, demonstrating reasoning skills in defined areas of the game. The next stage of agent-based systems applies these skills to real-world contexts.
Improving memory. They should remember the details that were said.
Better personalization. You need to understand your users’ preferences, likes and dislikes.
Use tools. They can use hardware such as robots from the physical world, software such as calculators, and even other AI systems.
However, Hassabis told the summit that he believes the full realization of artificial general intelligence, which can reason like humans, is still 10 years away.
Read the original article on Business Insider
